J. Am. Chem. Soc. 1998, 120, 1285-1288
1285
Photoisomerization Path for a Realistic Retinal Chromophore Model: The Nonatetraeniminium Cation Marco Garavelli,§ Thom Vreven,† Paolo Celani,† Fernando Bernardi,§ Michael A. Robb,*,† and Massimo Olivucci*,§ Contribution from the Dipartimento di Chimica “G.Ciamician” dell’ UniVersita` di Bologna, Via Selmi 2,40126 Bologna, Italy, and King’s College London, London WC2R 2LS, United Kingdom ReceiVed August 4, 1997. ReVised Manuscript ReceiVed NoVember 18, 1997
Abstract: In this paper, ab initio CASSCF computations are used to investigate the photoisomerization path of the protonated Schiff base (PSB) 4-cis-γ-methylnona-2,4,6,8-tetraeniminium cation: a five conjugated double bond model of the retinal chromophore of rhodopsin (the human retina visual pigment). We show that, after initial skeletal relaxation from the Franck-Condon region (which involves a large increase in the central CdC bond length), the system is “trapped” in an energy plateau on the S1 energy surface which may be the origin of the “slow” cis f trans isomerization dynamics observed in retinal PSBs in solution. The energy plateau is absent in shorter retinal chromophore models which have a steeper S1 isomerization path. The rhodopsin cavity (where the native chromophore is embedded) may have the effect of removing the energy plateau from the S1 potential thus dramatically increasing the photoisomerization rate from picoseconds to femtoseconds.
1. Introduction The 11-cis retinal protonated Schiff base (PSB11) is the chromophore of rhodopsin (the human retina visual pigment).
The photoisomerization of PSB11 to its all-trans isomer (PSBT) is one of the fastest chemical reactions observed so far.1 In particular, photoexcitation of PSB11 in rhodopsin yields a transient fluorescent state with a lifetime of 50-60 fs.2,3 After this state is left, ground-state PSBT is formed within 200 fs.4 In contrast, the photochemistry of free PSB11 in solution is 2 orders of magnitude slower: in methanol, PSB11 has a ca. 3 ps fluorescence lifetime and PSBT is formed in 10 ps.5 Similar excited-state lifetimes have been reported for the PSBT6,7 and 13-cis retinal protonated Schiff base (PSB13)6 chromophores. †
King’s College London. Universita` di Bologna. (1) Yoshizawa, T.; Kuwata, O. In Horspool, W. M., Song, P.-S., Eds. CRC Handbook of Organic Photochemistry and Photobiology; CRC Press: 1995; pp 1493-1499. (2) Kochendoerfer, G. G.; Mathies, R. A. J. Phys. Chem. 1996, 100, 14526-142532. (3) Kandori, H.; Sasabe, H.; Nakanishi, K.; Yoshizawa, T.; Mizukami, T.; Shichida, Y. J. Am. Chem. Soc. 1996, 118, 1002, 1005. (4) (a) Schoenlein, R. W.; Peteanu, L. A.; Mathies R. A.; Shank C. V. Science 1991, 254, 412-415. (b) Wang, Q.; Schoenlein, R. W.; Peteanu, L. A.; Mathies R. A.; Shank C. V. Science 1994, 266, 422-424. (5) (a) Kandori, H.; Katsuta, Y.; Ito, M.; Sasabe, H. J. Am. Chem. Soc. 1995, 117, 2669-2670. (6) Logunov, S. L.; Song, L.; El-Sayed, M. J. Phys. Chem. 1996, 100, 18586-18591. §
Recently we have investigated the structure of the S1 energy surface of the short PSB11 model 2-cis-penta-2,4-dieniminium cation (and its R-methylated form 1).8 A minimum energy path (MEP) computation shows that the relaxation on S1 occurs along a barrierless path which connects the Franck-Condon (FC) structure to a low-lying S1/S0 crossing. Formation of the ground state trans photoproduct occurs via decay at the crossing where the system has developed an ca. 80° twist about the central CdC bond. Semiclassical trajectory computations demonstrate that such energy surface leads to a 20-40 fs excited-state lifetime with photoproduct formation occurring on a ca. 100 fs time scale.9 Thus, the time scale for the transformation of the short model is much shorter of that of solution PSB11. In this paper, ab initio CASSCF computations with 10 electrons in 10 π-orbitals CAS and the 6-31G* basis set10 are used to investigate the structure of the S1 MEP of the longer PSB11 model 2 (4-cis-γ-methylnona-2,4,6,8-tetraeniminium (7) (a) Hamm, P.; Zurek, M.; Ro¨schinger, Patzelt, H.; Oesterhelt, D.; Zinth, W. Chem. Phys. Lett. 1996, 263, 613-621. (b) Kandori, H.; Sasabe, H. Chem. Phys. Lett. 1993, 216, 126-132. (8) Garavelli, M.; Celani, P.; Bernardi, F.; Robb, M. A.; Olivucci, M. J. Am. Chem. Soc. 1997, 119, 6891-6901. (9) Vreven, T.; Bernardi, F.; Garavelli, M.; Olivucci, M.; Robb, M. A.; Schlegel, H. B. J. Am. Chem. Soc. In press. (10) The MC-SCF program we used is implemented in Gaussian 94; Revision B.2; Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.; Peng, C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; HeadGordon, M.; Gonzalez, C.; Pople, J. A. Gaussian, Inc.: Pittsburgh, PA, 1995.
S0002-7863(97)02695-4 CCC: $15.00 © 1998 American Chemical Society Published on Web 02/03/1998
1286 J. Am. Chem. Soc., Vol. 120, No. 6, 1998 Table 1. Ground State Vibrational Frequencies (ωi) and Relative S1 Resonance Raman Intensities for the Totally Symmetric Modes of the CAS-SCF/6-31G* Optimized Geometry of 4-cis-γ-Methylnona-2,4,6,8-tetraeniminium Cationd ωi (cm-1)
relative intensities
expb
calca
expb
gradc
assignment
968 1005 1016 1190 1218 1237 1276 1439 1556 (br)
939 974 1044 1188 1262 1273 1311 1474 1580-16281630 1680
0.2 0.2 0.2 0.2 0.4 0.4 0.2 0.1 1.0
0.1 0.2 0.1 0.2 0.1 0.5 0.2 0.2 0.4-0.71.0 0.3
C-H rock C-H rock C-C stretch + C-H rock C-C stretch C-H bend
1658
0.1
C-C stretch + C-H bend C-H bend CdC stretch + CdN stretch CdC stretch
a Computed at the B3LYP/6-31G* level of theory. b Data for the protonated n-butylamine retinal (see ref 13); br ) broad. c The intensities are computed using the gradient formula in ref 12. d Notice that the most intense peak corresponds to the 1630 cm-1 stretch and that the shape of the spectra is reproduced.
cation). This is the first MEP computation attempted on a realistic retinal chromophore model. We show that, after initial skeletal relaxation from the FC point, 2 is “trapped” in an energy plateau (absent in model 1 where only an inflection region is detected) on the S1 energy surface. Model 2 does not include solvent, cavity or counterion, thus this finding suggests that the intramolecular force field of the PSB11 cation is at the origin of its slower excited-state dynamics. 2. Computational Methods MEP computations, in the full space of geometrical coordinates, were carried out using a combination of the IRD and IRC methods as already discussed in previous work.11 The MEP is computed by starting from the totally symmetric FC point of the system. Although accurate frequencies computations for model 2 are out of reach, the path has been determined by using a procedure which biases the search of energy valleys (see ref 8 for details). In particular, to avoid following totally symmetric ridges rather than real valleys, the MEP search was started using a fully asymmetric vector (generated via torsional deformation of the intersection structure) as the initial guess. The quality of our retinal protonated Schiff base model has been tested by computing the resonance Raman (RR) spectra12 and estimating the absorption and fluorescence λmax. Remarkably the computed frequencies and RR intensities (see Table 1) are in qualitative agreement with those observed for PSB11 in solution.13 Similarly the estimated14 S1-S0 energy gap at the FC structure and 1.5 au MEP structure yield absorption and fluorescence λmax values (428 and 618 nm) close to the (11) (a) Celani, P.; Robb, M. A.; Garavelli, M.; Bernardi F.; Olivucci M. Chem. Phys. Lett. 1995, 243, 1-8. (b) Garavelli, M.; Celani, P.; Fato, M.; Bearpark, M. J.; Smith, B. R.; Olivucci, M.; Robb, M. A. J. Phys. Chem. A 1997, 101, 2023-2032. (12) Orlandi, G.; Zerbetto, F.; Zgierski, M. Z. Chem. ReV. (Washington, D.C.) 1991, 91, 867. A complete analysis of RR spectra of various retinal PSB models will be published separately. (13) Mathies, R.; Freedman, T. B.; Stryer, L. J. Mol. Biol. 1997, 109, 367-372. (14) The dynamic correlation energy correction to the CASSCF energy gap is conveniently evaluated using the PT2F method available in MOLCAS (MOLCAS, Version 3; Andersson, K.; Blomberg, M. R. A.; Fu¨lscher, M.; Kello¨, V.; Lindh, R.; Malmqvist, P.-A.; Noga, J.; Olsen, J.; Roos, B. O.; Sadlej, A. J.; Siegbahn, P. E. M.; Urban, M.; Widmark, P. O. University of Lund, Sweden 1994). However, compound 2 is to large for such a calculation. Thus we estimate the magnitude of the correction by performing PT2F computations at the FC and S1 minimum (planarity constrained) points of the lower homologue all-trans hepta-2,4,6-trieniminium cation. This yield a -15.6 and -8.2 kcal mol-1 correction for the energy gap at FC and plateau region (MEP point at 1.5 au), respectively.
GaraVelli et al. values observed for PSB11 and PSBT in hexane (457 and 620 nm)15 or methanol (442 and 675).5,15 These data suggest that the general structure of the energy surface of the isolated cation 2 may be similar to that of PSB11 in solution. This view is also supported by the fact that the absorption and fluorescence λmax for PSBT in an apolar environment (hexane, hexadecane) are only ca. 10 nm higher and 50 nm lower that in methanol15 (which indicate energy differences
